Search

Few HMD-based virtual environment systems display a rendering of the user's own body. Subjectively, this often leads to a sense of disembodiment in the virtual world. We explore the effect of being able to see one's own body in such systems on an objective measure of the accuracy of one form of space perception. Using an action-based response measure, we found that participants who explored near space while seeing a fully-articulated and tracked visual representation of themselves subsequently made more accurate judgments of absolute egocentric distance to locations ranging from 4 m to 6 m away from where they were standing than did participants who saw no avatar. A nonanimated avatar also improved distance judgments, but by a lesser amount. Participants who viewed either animated or static avatars positioned 3 m in front of their own position made subsequent distance judgments with similar accuracy to the participants who viewed the equivalent animated or static avatar positioned at their own location. We discuss the implications of these results on theories of embodied perception in virtual environments. [ABSTRACT FROM AUTHOR]

Mental rotation is the capacity to predict the orientation of an object or the layout of a scene after a change in viewpoint. Previous studies have shown that the cognitive cost of mental rotations is reduced when the viewpoint change results from the observer's motion rather than the object or spatial layout's rotation. The classical interpretation for these findings involves the use of automatic updating mechanisms triggered during self-motion. Nevertheless, little is known about how this process is triggered and particularly how sensory cues combine in order to facilitate mental rotations. The previously existing setups, either real or virtual, did not allow disentangling the different sensory contributions, which motivated the development of a new high-end virtual reality platform overcoming these technical limitations. In the present paper we will start by a didactic review of the literature on mental rotations and expose the current technical limitations. Then we will fully describe the experimental platform that was developed at the Max Planck Institute for Biological Cybernetics in Tübingen. The setup consisted of a cabin mounted on the top of a six degree-of-freedom Stewart platform inside of which was an adjustable seat, a physical table with a screen embedded, and a large projection screen. A 5-PC cluster running Virtools was used to drive the platform and render the two passive stereovision scenes that were displayed on the table and background screens. Finally, we will present the experiment using this setup that allowed replicating the classical advantage found for a moving observer, which validates our setup. We will conclude by discussing the experimental validation and the advantages of such a setup. [ABSTRACT FROM AUTHOR]

We attempt to shed light on the algorithms humans use to classify images of human faces according to their gender. For this, a novel methodology combining human psychophysics and machine learning is introduced. We proceed as follows. First, we apply principal component analysis (PCA) on the pixel information of the face stimuli. We then obtain a data set composed of these PCA eigenvectors combined with the subjects' gender estimates of the corresponding stimuli. Second, we model the gender classification process on this data set using a separating hyperplane (SH) between both classes. This SH is computed using algorithms from machine learning: the support vector machine (SVM), the relevance vector machine, the prototype classifier, and the K-means classifier. The classification behavior of humans and machines is then analyzed in three steps. First, the classification errors of humans and machines are compared for the various classifiers, and we also assess how well machines can recreate the subjects' internal decision boundary by studying the training errors of the machines. Second, we study the correlations between the rank-order of the subjects' responses to each stimulus—the gender estimate with its reaction time and confidence rating—and the rank-order of the distance of these stimuli to the SH. Finally, we attempt to compare the metric of the representations used by humans and machines for classification by relating the subjects' gender estimate of each stimulus and the distance of this stimulus to the SH. While we show that the classification error alone is not a sufficient selection criterion between the different algorithms humans might use to classify face stimuli, the distance of these stimuli to the SH is shown to capture essentials of the internal decision space of humans. Furthermore, algorithms such as the prototype classifier using stimuli in the center of the classes are shown to be less adapted to model human classification behavior than algorithms such as the SVM based on stimuli close to the boundary between the classes. [ABSTRACT FROM AUTHOR]

The literature often suggests that proprioceptive and especially vestibular cues are required for navigation and spatial orientation tasks involving rotations of the observer. To test this notion, we conducted a set of experiments in virtual environments in which only visual cues were provided. Participants had to execute turns, reproduce distances, or perform triangle completion tasks. Most experiments were performed in a simulated 3D field of blobs, thus restricting navigation strategies to path integration based on optic flow. For our experimental set-up (half-cylindrical 180 deg. projection screen), optic flow information alone proved to be sufficient for untrained participants to perform turns and reproduce distances with negligible systematic errors, irrespective of movement velocity. Path integration by optic flow was sufficient for homing by triangle completion, but homing distances were biased towards the mean response. Additional landmarks that were only temporarily available did not improve homing performance. However, navigation by stable, reliable landmarks led to almost perfect homing performance. Mental spatial ability test scores correlated positively with homing performance, especially for the more complex triangle completion tasks—suggesting that mental spatial abilities might be a determining factor for navigation performance. In summary, visual path integration without any vestibular or kinesthetic cues can be sufficient for elementary navigation tasks like rotations, translations, and triangle completion. [ABSTRACT FROM AUTHOR]